data fusion for precision turf managementprecision turf management •precision turf treatment...

DATA FUSION FOR PRECISION TURF MANAGEMENT

P r e s e nte d b y K i r k M . S t u e v e

A s s i s t a n t P r o f e s s o r o f G e o s c i e n c e s a t M i n n e s o t a S t a t e U n i v e r s i t y M o o r h e a d

S e n i o r G I S C o n s u l t a n t a t F r o s t I n c .

TA B L E O F C O N T E N T S

• Overview of precision turf management

• Overview of data fusion

• Demonstration of data fusion• High-resolution WorldView2 satellite data• NDVI and NDRE turf health indices

• Recommended protocols

• Questions and comments

P R E C I S I O N T U R F M A N A G E M E N T

• In a perfect world:

P R E C I S I O N T U R F M A N A G E M E N T

• In a reality:

P R E C I S I O N T U R F M A N A G E M E N T

• Precision turf treatment:

P R E C I S I O N T U R F M A N A G E M E N T

• Precision turf treatment pitfalls:

• Snapshot from one data source frequently used

• No satellite, camera, or other sensor can fully capture your turf conditions

• When comparing data, different patterns and trends often emerge

• NDVI and NDRE example from WorldView2 satellite below:• Almost 100,000 SQFT difference on 18-hole course with over 1 million SQFT

• Which one is right? What should one do?

Normalized Difference Vegetation Index (NDVI) Normalized Difference Red Edge Index (NDRE)

D ATA F U S I O N & P R E C I S I O N T U R F

• Recognizes 3 laws of ecology:• “Too much” and “too little” can stress or kill turf

• Exact thresholds of “too much” and “too little” may vary on individual holes or between different holes and depend on “many factors interacting together”

• Data fusion:• All types of data have limitations

• Can benefit from combining data and calculations

• Create single precision management layer from data combinations that is actionable

Turf Health Index 1

Turf Health Index 2

Soils

Hydrology

Topography

ONE ACTIONABLE RX

D ATA F U S I O N E X A M P L E

• Fairways Demonstration• Fuse two key turf health indices from

satellite

• 2-m WorldView-2 satellite data

• NDVI and NDRE turf health indices from the July 10 of the 2012 growing season

• 18 holes on a Minnesota golf course (6 holes displayed for viewing convenience)

• Management Problem• Targeted application of fungicide to

lushest/greenest parts of fairway

• Fungi most likely to establish in these area; want to apply before visual confirmation of establishment

D ATA F U S I O N & P R E C I S I O N T U R F

NDVI (7-5/7+5, 0.331-0.832) NDRE (6-5/6+5, 0.423-0.796)

D ATA F U S I O N E X A M P L E

NDVI (Normalize, make zones) NDRE (Normalize, make zones)

D ATA F U S I O N E X A M P L E

NDVI (152,000 SQFT LUSH TURF) NDRE (57,000 SQFT LUSH TURF)

D ATA F U S I O N E X A M P L E

NDVI LAYER ON TOP OF NDRE NDRE LAYER ON TO OF NDVI

D ATA F U S I O N E X A M P L E

NDVI AND NDRE ANALYTICS: DEVELOPING A PLAN

0

20

40

60

80

100

120

140

160

NDVI = NDRE NDVI NDVI + NDRE NDRE +

49.25

152.07

102.82

57.04

7.79

SQ

FT

(T

HO

US

AN

DS

)

NDVI & NDRE Analytics

D ATA F U S I O N E X A M P L E

ACTION RX 1: Include all areas mapped by NDVI and NDRE (160,000 SQFT)

*AGGRESSIVE!

D ATA F U S I O N E X A M P L E

ACTION RX 2: Only include areas where NDVI and NDRE overlap (50,000 SQFT)

*CONSERVATIVE!

D ATA F U S I O N E X A M P L E

CAUTION: - Remote sensing problems on golf courses

- Trees and other obstructions

TREES

I N I T I A L P R OTO C O L S

• (1) Identify a management problem that you want to address on your course

• (2) Don’t rely on a single piece of information to make an RX for solving the problem

• (3) Use on-the-ground knowledge of your course to “mine” multiple data sources that might help address your management problem

• (4) Let your knowledge of the course guide the data fusion process (i.e., integrate the most useful parcels of information from different data sources into ONE actionable RX file)

• (5) Manipulate data so it compliments your precision equipment

S O F T WA R E U S E D F O R A N A LY S E S • All open source (i.e., free) software was used for the remote sensing and GIS analyses

of the data presented today:

• QGIS: http://www.qgis.org/en/site/

• R: https://www.r-project.org/

• OSGeo: http://www.osgeo.org/

• GDAL: http://www.gdal.org/

• FWTools: http://fwtools.maptools.org/

F U T U R E W O R K

• Partners

• Didn’t have enough time or resources to collect and integrate additional data• e.g., soil moisture, compaction, and nutrient data

• Looking for partners to collaborate

• Contact Kirk after the talk or at 651-491-3372 or kirk@frostserv.com if you’re interested

Q U ES T I O N S ? ?